Back to previous post: Bloomsday

Go to Making Light's front page.

Forward to next post: Typesetting: when it changed

Subscribe (via RSS) to this post's comment thread. (What does this mean? Here's a quick introduction.)

June 16, 2004

Hot jets!
Posted by Teresa at 03:38 PM *

You know that old scientifictional idea about a device (usually a cloak) that renders you invisible by projecting an image onto its surface of whatever’s behind you? Professor Susumu Tachi says he’s done it. He has the technology. And what form has it taken? Why, a cloak, of course. He’s even given his material a great skiffy name: retro-reflectum. (via Laura Mixon)

Comments on Hot jets!:
#1 ::: Eric Sadoyama ::: (view all by) ::: June 16, 2004, 03:50 PM:

At the prof's university website, this brochure has a few neat examples of practical uses. I like the one where the automobile driver can "see through" the rear of the car when reversing into the garage, although the one where airline pilots can see the runway beneath their cockpit floor would probably have much more impact.

#2 ::: Will "scifantasy" Frank ::: (view all by) ::: June 16, 2004, 04:03 PM:

Great idea as it is, and I recall hearing about it some months ago, I had to crack up at this:

"I wanted to create a vision of invisibility."

Um...

#3 ::: Xopher ::: (view all by) ::: June 16, 2004, 04:06 PM:

So that's what an invisible barrier looks like!!!

That said, I want one. Want, want, want.

#4 ::: Michael ::: (view all by) ::: June 16, 2004, 04:07 PM:

Want. One.

I have to say, however, that I'm not sure I wouldn't find seeing through certain things (walls, floors, cars, etc.) a bit distressing. And probably painful, as I'd surely be running into them pretty often.

#5 ::: James Angove ::: (view all by) ::: June 16, 2004, 04:13 PM:

Eric: The impact on aircraft will probably be minor, actually. Being able to figure out where the ground is is largely a solved problem for aircraft. At this point, computers will pretty much tell you where the glide path is.

On the other hand, as an adolecent, I knocked my folks garage door off its rails twice (and, wonder of wonders, managed to hide this fact. So don't call tell them.)

#6 ::: Xopher ::: (view all by) ::: June 16, 2004, 04:14 PM:

I assume the two sides don't have to be adjacent? And that one side could be opaque? (You might want to be able to see OUT of your house without other people seeing IN.)

#7 ::: Andy Perrin ::: (view all by) ::: June 16, 2004, 04:19 PM:

I can see using this method of invisibility for drivers and pirates -er pilots, but for military purposes more development is needed. What happens if you're hiding from the enemy, and the battery dies?
---------

My favorite piece of new tech would have to be the electronic ink reader developed by e-ink. This thing could eliminate my textbooks. But before I buy, I want a bigger one that can take any PDF I care to throw on it, not just some proprietary format.

#8 ::: Adam ::: (view all by) ::: June 16, 2004, 04:21 PM:

Looks like we can finally start the Panther Moderns now!

Where 'dat Steppin' Razor?

#9 ::: John M. Ford ::: (view all by) ::: June 16, 2004, 04:49 PM:

Hey . . . a devicing cloak.

Never mind.

A couple of years ago, there was a brief paper referenced in NASA TECH BRIEFS (thanks again, Paula) about using cameras and projection to Hide Stuff (the example given was a tank). Not that that's prior art or anything -- the idea is not difficult to come up with, the implementation is.

"Did you hear something?"

"Well, sure. Sounded like a pile of cheap glassware falling half a ton of silverware. But I don't see a darn thing."

"Okay. Nothin' to worry about, then. Tell you what, let's go hang out near the carefully hidden thermynucular interociter, since nobody else is here."

"Yeah. The geeks hide all their beer in the supercondunking thingummy. Will you quit singing 'Bad Moon Rising'?"

"I'm not singing. You quit singing."

#10 ::: David Moles ::: (view all by) ::: June 16, 2004, 04:51 PM:

“However there are massive questions of potential misuse too, particularly surrounding the huge crime implications. It would become incredibly difficult to spot a thief, for example, if the items they were taking were simply disappearing under the cloak.”

Uh, yeah, and this is also why those strange new gadgets called "coats" have caused such a spike in crime rates . . .

#11 ::: Jon H ::: (view all by) ::: June 16, 2004, 05:03 PM:

The invisibility cloak application of this reminds me of Al Franken's mobile satellite uplink getup, where he wore a satellite dish on his head.

In this case, the helmet would have an arm coming out front, holding a projector aimed at the wearer, and a camera on the back, looking behind the wearer.

Frankly, the whole idea sounds like it was conceived by some pot-addled undergrads in a dorm room at 4:30 am, playing with a projector screen they stole from a classroom.

"Look, I'm invisible! Nobody's here but this shadow bunny on my shirt."

#12 ::: Moira ::: (view all by) ::: June 16, 2004, 05:04 PM:

retro-reflectum

That....is just an entirely unfortunate combination of syllables.

#13 ::: Patrick Connors ::: (view all by) ::: June 16, 2004, 05:04 PM:

Not to burst anyone's bubble - we could all use an invisibility cloak at times - but what you don't see there is the camera to take the picture of what's behind the cloak, nor the projector that's needed to put the image onto the cloak.

Special effects geeks who look up the details will recognize this as a modification of front-screen projection. This works best from one single point.

Cameras are pretty small these days, but a good projector will be hard to hide.

- sigh -

#14 ::: Jon H ::: (view all by) ::: June 16, 2004, 05:10 PM:

Actually, I can imagine one application: it would probably work as a way of fooling a security camera, a high-tech version of the "take a polaroid picture and stick it on the lens" approach.

Seen through a lo-fi security camera, it'd probably be quite convincing. Not the best way of going about it.

The whole issue of needing a projector really kills the whole idea. It'd be much better if it was some kind of digital paper-like technology, which could create images within itself.

The "see-through airplane cockpit" idea would be much more feasible if it didn't need a projector, but instead consisted of an active digital-paper coating applied to the surfaces of the cockpit.

#15 ::: Andy Perrin ::: (view all by) ::: June 16, 2004, 05:19 PM:

Jon H suggested: The "see-through airplane cockpit" idea would be much more feasible if it didn't need a projector, but instead consisted of an active digital-paper coating applied to the surfaces of the cockpit.

I'm thinking along the same lines (see my e-ink post above). To handle multiple viewpoints, you could have a screen that works like those gimmicky ads that switch between two or more (still) pictures as you look from side to side. Apple has a neat one that shows a picture of the new G5 from the outside when viewed straight on, and lets you see through to the inside when you look at it obliquely.

#16 ::: Claude Muncey ::: (view all by) ::: June 16, 2004, 05:46 PM:

Beat me to it, Patrick --

This is real-time digital front projection. The basic technique has been around in special effects for a century or more but was perfected by Kubrick, Pederson, Trumbull and others for 2001 (check out this article from American Cinematographer about it). What the good professor seems to have come up with is a combination of digital front projection technology combined with a more fabric-like version of 3M ScotchLite which is one of the materials used for front-projection screens. Of course I don't know the limits of his materials, but ScotchLite and it's competitors have a rather narrow effective cone of reflection, which is side-effect of their powerful retro-reflective capabilities. As 3M sells tons of this stuff, I hope this professor has done a good patent search.

The practical problem with this technology is that the camera (which these days could be quite small), the projector and the viewer must be aligned along the same optical axis, with little tolerance for alignment problems.

One idea I had years ago for a story (which was never written of course) is vehicle with a flat window, slated toward the driver to act as a reflector for a (yet to be developed) digital holgraphic display. The information from the display would be seen as being "projected" onto objects out in front of the vehicle, and being holographic, whould not have an alignment problem -- you could move your head around as much as you wanted. Consider projecting 3D radar results this way in fog or at night without using headlights, combining overhead imaging (sattelite pictures), sensor overlays, GPS driven navigational information, targeting for your weapons and such. Sort of combining a window with VR.

Now all I need is holographic technology that may never be invented. Sigh.

#17 ::: Claude Muncey ::: (view all by) ::: June 16, 2004, 05:48 PM:

Of course the holographic technology may arrive before I learn to spell . . .

#18 ::: Jill Smith ::: (view all by) ::: June 16, 2004, 05:54 PM:

Tangentially, an industrial designer friend of mine led me to this article about concrete laced with optical fibers - you can see shadows right through a concrete wall using this stuff. Pretty cool.

#19 ::: Charlie Stross ::: (view all by) ::: June 16, 2004, 06:00 PM:

Hmm. If I was placing bets on this in a military context, I'd plonk my money down on hyperspectral imagers.

Human colour vision distinguishes red/green/blue because those are the wavelengths the pigments in our retinal cone cells absorb. But CCDs aren't limited to the same broad smears of wavelengths. We can build sensors that are much more sensitive to colour gradations than the human eye, and something that is camouflaged to the point of invisibility to the Mk.1 eyeball stands out like a sore thumb to a hyperspectral imager.

So. Who's willing to bet me that this gizmo is designed to fool the human eye, but shows up like a beacon in far infra-red or near ultra-violet, never mind some weird subtle hues of greenish-grey?

#20 ::: John M. Ford ::: (view all by) ::: June 16, 2004, 06:22 PM:

". . . digital front projection . . ."

A digital extension of the Jenkins Process, named for its inventor, Will F. Jenkins, aka Murray Leinster. He wrote an account of the invention decades back in ANALOG. One of the observations I recall was that the movie business was perfectly willing to use the idea (it was far superior to rear projection) . . . but wasn't very interested in paying the inventor anything for using it.

#21 ::: Mary Kay ::: (view all by) ::: June 16, 2004, 06:40 PM:

Gosh. Y'all guys ought to write that Sci Fi stuff!

MKK

#22 ::: Jon H ::: (view all by) ::: June 16, 2004, 06:51 PM:

I wonder what happens if you view the image through a polarized lens rotated to various orientations.

If the reflected light of the projection is polarized, it would likely disappear or fade when the lens is turned to some angle.

#23 ::: Julia Jones ::: (view all by) ::: June 16, 2004, 07:03 PM:

Charlie said: Human colour vision distinguishes red/green/blue because those are the wavelengths the pigments in our retinal cone cells absorb. But CCDs aren't limited to the same broad smears of wavelengths. We can build sensors that are much more sensitive to colour gradations than the human eye, and something that is camouflaged to the point of invisibility to the Mk.1 eyeball stands out like a sore thumb to a hyperspectral imager.

It might also stand out like a sore thumb to anyone who doesn't have standard colour vision. I wonder if he's tested it on people with red/green colour blindness, or any of the rarer ones?

#24 ::: Claude Muncey ::: (view all by) ::: June 16, 2004, 07:21 PM:

John, this short history of ScotchLite and front projection technology (from a 1984 SMPTE Conference), seems to give Jenkins deserved credit. Unfortunately, it also indicates he had problems with the patents:

There were numerous advantages to this approach versus Thorner's mirror, and these were recognized not only by Meyer but also by Will Jenkins. In 1952 and in 1953, Jenkins filed two patents detailing a plethora of different methods utilizing Scotchlite as a catadioptric front projection system. Jenkins was followed in 1962 by two French cinematographers, Alekan and Gerard, who were somehow able to obtain a British patent for a process that, aside from a change in the refractive index of the glass from which the material was fabricated, was identical to Jenkins'.

One additional observation about this new appplication of front projection is that it works well in a dark room, but not nearly as well outside during the day due to the higher ambient light level.

#25 ::: Claude Muncey ::: (view all by) ::: June 16, 2004, 07:26 PM:

Actually, Jon, if you want, you could project images from two different cameras, each projected through a polarizing filter, one rotated 90 degrees from the other. With the right glasses, you get a 3D effect. This works well with this kind of retroreflective surface as each image would be reflected back independently. With two properly aligned cameras, properly set up projectors, and just a little digital processing, you would get an even more powerful "invisibility" effect.

#26 ::: Michelle ::: (view all by) ::: June 16, 2004, 07:40 PM:

I knew I'd seen this before.
http://news.bbc.co.uk/2/hi/asia-pacific/2777111.stm

He's been pushing it for awhile.

But that doesn't make it any less neat.

#27 ::: Virge ::: (view all by) ::: June 16, 2004, 08:42 PM:

Face it. The cloak is a media attention grabber to promote the "see through walls" technology. The "invisible" cloak would be painfully obvious to any observer.
1. Angle of view - trompe l'oeil is very limited.
2. Projector placement - did you want to be able to move while you're wearing it?
3. Dynamic range - the range of light intensities in an outdoor setting is far wider than current projectors can produce.

The reporter extrapolates:
"There are many potential uses of the cloak, ranging from espionage and military purposes"
and
"It would become incredibly difficult to spot a thief, for example, if the items they were taking were simply disappearing under the cloak."

What can I say? Too much Harry Potter.
If the implementation lived up to the concept it would be really neat. This is just a cheap publicity grab.

#28 ::: Claude Muncey ::: (view all by) ::: June 17, 2004, 12:50 PM:

Now all I need is holographic technology that may never be invented.

Maybe I spoke too soon . . .

#29 ::: Dolloch ::: (view all by) ::: June 17, 2004, 01:26 PM:

To answer an earlier question, I'd say the cone of reflection is pretty wide (long? big? circumference-licious?). If you follow the outside link on the BBC site, you can see the quicktime movies of its application. The ball and brick are probably the most impressive. It seems that the ambient light shadows on the object itself cause a lot of the invisibility failure.

In any case - amazing stuff! Imagine - movie theatres that don't need permanant screen structures to hold the screen absolutely flat. Hand out headphones and you could just lower a screen in the middle of the theatre and get 2 theatres without construction! Perfect for small art films that don't have a big audience. Or, portable drive-ins anyone? No more student film festivals projected on sheets on the side of the Quad building! Home use - big screen tv wall projection. Or, with a webcam-like device, a picture of your backyard while your children play.

Very interesting. Er, is he selling stock...?

#30 ::: Dave Bell ::: (view all by) ::: June 17, 2004, 04:48 PM:

Colour matching (and hyperspectral effects) do matter for camouflage, but they're not the only elements. You don't have to match either the precise colour or the shape of the leaves behind you to look like something with leaves.

Infra-red already has to be taken into account. Green leaves reflect a lot. Most green paints don't. And seeing infra-red on the battlefield is an old technology.

#32 ::: Dan Boone ::: (view all by) ::: June 18, 2004, 05:11 PM:

"What happens if you're hiding from the enemy, and the battery dies?"

Uh, er, roughtly the same thing that happens when you are flying along in a fighter jet and you run out of fuel?

I'm sure there are lots of perfectly good practicality objections to this thing. But that's not one of them. The world is full of marvelously useful devices that can be fatal to the user when vital consumables are exhausted.

#33 ::: Andy Perrin ::: (view all by) ::: June 18, 2004, 05:34 PM:

Dan Boone said:
"What happens if you're hiding from the enemy, and the battery dies?"

Uh, er, roughtly the same thing that happens when you are flying along in a fighter jet and you run out of fuel?

If the power failed on this thing, then the user would be instantly exposed. It would be nice if the material retained the last image projected on it even after the power is cut. An e-ink type system would have this property. As for fighter jets, couldn't they glide for a short time?

#34 ::: Graydon ::: (view all by) ::: June 18, 2004, 11:32 PM:

Andy --

You might find the story about Juvat Bingo Fuel of interest; many modern fighters are not so good at gliding without power.

#35 ::: Andy Perrin ::: (view all by) ::: June 19, 2004, 12:19 AM:

Graydon, that link proves my point. He made it down; the plane did not go thud.

#36 ::: Graydon ::: (view all by) ::: June 19, 2004, 12:29 AM:

Andy --

The plane did not go thud by a combination of great good luck and auxillary power provided via hydrazine. This is a big deal because if an F16 loses power, it becomes aerodynamically unstable and promptly crashes.

#37 ::: Andy Perrin ::: (view all by) ::: June 19, 2004, 12:43 AM:

The point is that you want a design to be somewhat forgiving if possible; the fact that the F-16 is not statically stable is a point against it in this sense. It's a design trade-off to improve the plane's ability to maneuver. In the case of the camouflage, there isn't anything gained by having it constantly drain power to maintain the image.

#38 ::: John M. Ford ::: (view all by) ::: June 19, 2004, 12:57 AM:

Well, Andy, no, it doesn't; he landed on the APU (Auxiliary Power Unit) with two minutes to spare. Modern jet fighters' wings are primarily control surfaces; without thrust, they will not keep the bird airborne.

The question is about failure modes. If your fighter is no longer producing thrust -- which means by all means, including fan inertia and the APU -- you're going to faw down go boom, and you should perhaps consider the Martin Baker device under your seat, which is usable for brief levitation. It even comes with a free necktie. (For those who missed the nuance, "bingo fuel" means "absolute minimum fuel to reach your intended destination;" "Juvat bingo fuel" is an in-joke derived from that.)

If the Tarnmäntel stops working, obviously you aren't invisible anymore. Now, sensibly, one would never rely 100% on this thing, just as one never sensibly relies 100% on any gadget, even if you're not in a war zone. Conventional hiding and sneaking plus Rivendell Poncho will get you farther than cloak alone. (Maybe they'll add the fabulous 180° audio phase silencer.) It's difficult to conceive of anybody except Rumsfeld imagining that you could put a full battalion in these things and have them move cross-country unseen, unheard, and unsmelled by the bunny rabbits and the majestic moose; it's a device for no more than a squad on a specific in-and-out task, which is going to take as little time as the people doing it can possibly manage. On that basis, it could work. If it worked in the first place. If we had some ham, we could go to Mars, if we had a workable interplanetary flight system.

I am suddenly reminded of a double-page spread in Popular Science about forty years ago. Bell had just shown off the first version of the Bell Rocket Belt, and this illo showed flying GIs blasting Commie tanks with bazookas. The idea of firing a bazooka while wearing a literal buttload of propellant may have been the single weirdest thing about it.

Next up on Wonderful New Military Advances: DoD to fund research into Jedi mind tricks. "These aren't the incriminating memos you're looking for."

#39 ::: Andy Perrin ::: (view all by) ::: June 19, 2004, 01:17 AM:

Mr. Ford said (and BTW, do you prefer 'Mike' or 'John' from acquaintances?):

Well, Andy, no, it doesn't; he landed on the APU (Auxiliary Power Unit) with two minutes to spare. Modern jet fighters' wings are primarily control surfaces; without thrust, they will not keep the bird airborne.

Yup.

The question is about failure modes. If your fighter is no longer producing thrust -- which means by all means, including fan inertia and the APU -- you're going to faw down go boom...

Agreed. In my book, this is not in their favor. But there's not much to be done about it if you want to preserve maneuverability.

If the Tarnmäntel stops working, obviously you aren't invisible anymore. Now, sensibly, one would never rely 100% on this thing, just as one never sensibly relies 100% on any gadget, even if you're not in a war zone. Conventional hiding and sneaking plus Rivendell Poncho will get you farther than cloak alone.

A backup is always a good idea. I just think that if it is possible to design something in a manner that will fail gracefully, you should do it.

#40 ::: Randolph Fritz ::: (view all by) ::: June 19, 2004, 04:51 AM:

Oh, and here's the tasp.

#41 ::: David Goldfarb ::: (view all by) ::: June 20, 2004, 05:39 PM:

That's not a tasp, it's a droud. I'd imagine a tasp would be much harder....

#42 ::: Paula Lieberman ::: (view all by) ::: June 27, 2004, 11:41 AM:

"So. Who's willing to bet me that this gizmo is designed to fool the human eye, but shows up like a beacon in far infra-red or near ultra-violet, never mind some weird subtle hues of greenish-grey?"

No bet from over here.

Fooling sensors in one range of electromagnetic frequencies is one thing, fooling sensors from "DC to daylight" [actually, DC to gamma rays...] is quite another. It was something I thought a lot about during my defense industry years. If you want to hide a space vehicle in space you either need to have it not be detected as an Object of Interest, either by having it not show up in a whatever frequency bands an observer might have detectors in pointed in its direction, or buy having it look like something uninteresting to ignore that's not going to be like "hey, why is there something like looks like a pigeon flying around in the Class 10 Clean Room?!"

#43 ::: Bill Blum ::: (view all by) ::: June 27, 2004, 02:07 PM:

Fooling sensors in one range of electromagnetic frequencies is one thing, fooling sensors from "DC to daylight" [actually, DC to gamma rays...] is quite another.

The other problem with that is, sometimes you run into the problem of falling below the background noise--- "what's that airplane sized hole doing in my background sky temperature?"

#44 ::: Paula Lieberman ::: (view all by) ::: July 05, 2004, 05:48 PM:

Bill Blum wrote,

"Fooling sensors in one range of electromagnetic frequencies is one thing, fooling sensors from "DC to daylight" [actually, DC to gamma rays...] is quite another."

"The other problem with that is, sometimes you run into the problem of falling below the background noise--- "what's that airplane sized hole doing in my background sky temperature?""

That's a subset of "does not fool sensors."

'"What was the strange thing the dog did in the night?"'

Dog, -what- dog?!

"This is what the background noise typically looks like. This does -not- look like background noise, it does not have the appropriate level of white or pink noise in it. Something is -wrong-.

Sort of like ways of telling the communications link is down. Some of them are set up so that there is a constant stream of signal going through, if the signal isn't there, the comm link's gone down. That serves other purposed, too. Trying to dig signal out of intentional continuing signal that's really -noise- is more difficult than "hey, look, there's a message!"

It's one of the ways that someone can discourages eavesdroppers -- bore them to death with chatter that's utterly meaningless such that they decide either there's nothing worth listening to, or it's too much effort to sort through all the chatter looking for -information-. However, if there an easy way to discriminate such as tone of voice, or person turning an interesting color, or showing some nervous habit...

Doctored digital images tend to show up as being doctored, because they don;'t have the same level of statistical randomness as undoctored "natural" imagery. PhotoShop pictures that get processed too much tend to show obvious unnatural moire patterns, for example. Doing various types of transforms, doctors images show characters in the transform space that practically scream "doctored image!"

The same is true for things like sky surveys....

Choose:
Smaller type (our default)
Larger type
Even larger type, with serifs

Dire legal notice
Making Light copyright 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020 by Patrick & Teresa Nielsen Hayden. All rights reserved.